International comparison of fossil power efficiency and CO2 intensity - Update 2017 FINAL REPORT

Similar documents
International comparison of fossil power efficiency and CO 2 intensity - Update 2013 FINAL REPORT

Energy and CO 2 emissions in the OECD

Electricity and heat statistics

Power Generation: A Continued Role for Coal?

Background paper. Electricity production from wind and solar photovoltaic power in the EU

Competitive energy landscape in Europe

Energy, Electricity and Nuclear Power Estimates for the Period up to 2050

statistics Electricity information: Overview

WIND POWER TARGETS FOR EUROPE: 75,000 MW by 2010

Annual Electricity and Heat Questionnaire Overview Energy Statistics Training Paris, 4-8 March, 2013

Figure ES.1: LCOE ranges for baseload technologies (at each discount rate) 3% 7% 10%

Data, tables, statistics and maps ENERGY STATISTICS

Benchmarking Paper Mill Energy Efficiency and GHG Emissions for Major Producer Countries

Energy Efficiency Indicators: The Electric Power Sector

Highlights. Figure 1. World Marketed Energy Consumption by Region,

ITAM Mexico City, 27 October 2016

Please note that all IEA data is subject to the following Terms and Conditions found on the IEA s website:

ISSN energy. in figures. Energy

Present and future of Renewable Energies in Europe

IEA data collection on RES

Har du brug for flere data? 2006 Energistatistik 2006 Data Kort

THE PROSPECTS FOR RENEWABLES IN THE EU

Long-term Market Analysis Nordics and Europe Executive summary

Chapter 2. Case Studies: Keeping CO 2 emission at 2013 level by March 2018

energy in figures Energy

Status and Prospects of PV Technology

International Energy Agency Biofuels & Bioenergy Technology Roadmaps

Contribution of Renewables to Energy Security Cédric PHILIBERT Renewable Energy Division

Biomass Cogeneration Network- BIOCOGEN

International Index of Energy Security Risk

Strong focus on market and policy analysis

Energy Scenario for Achieving Sustainability in Indian Situations

Quick overview. Energy balance Production of primary energy. Renewable energy. Electricity and district heating. Consumption review 2009

The Danish Experience and View on Energy System Development

THE RENEWABLE ENERGY CONTRACTOR MARKET IN EUROPE THE POWER INDEX

Figure 1. Energy dependency of EU member states in 2004 (%)

EU wide energy scenarios until 2050 generated with the TIMES model

WORLDWIDE ENERGY SCENARIO AN OVERVIEW

Electricity and Heat. HP Chung UNESCAP Statistics Division. Workshop on Energy Statistics for ASEAN Countries

A decade of oil demand

Consequences of the German Energy Transition on the operation regime and the availability of coal fired power plants

Environmental Action Plan By the Japanese Electric Utility Industry

Global Energy Production & Use 101

Wind energy and Climate policy Fixing the Emission Trading System

COMPETITIVENESS OF NUCLEAR ENERGY AN INTERNATIONAL VIEWPOINT. Nuclear Power Plants for Poland, Warsaw 1-2 June 2006

Energy in Sweden 2018 An overview

Summary The world's capacity of renewable energies including hydropower is expected to increase from 2,200 GW at the end of 2017 to over 2,600 GW at t

JAPAN S ENERGY POLICY AND JAPAN-RUSSIA ENERGY COOPERATION

Milken Institute: Center for Accelerating Energy Solutions

Power market integration, focusing on the CHP switch to biomass and electric heating to replace fossil-fuelled heating

Potential of Solar Thermal in Europe. Werner Weiss, AEE Institute for Sustainable Technologies Peter Biermayr, Vienna University of Technology

Energy Efficiency Are we fighting windmills?

Medium Term Renewable Energy Market Report Michael Waldron Senior Energy Market Analyst Renewable Energy Division International Energy Agency

Photo: Thinkstock. Wind in power 2010 European statistics. February The European Wind energy association

Introduction: The Growing Importance of Biofuels in Asia

Medium Term Renewable Energy Market Report 2013

WIND ENERGY - THE FACTS PART VI SCENARIOS AND TARGETS

Lecture 1: Energy and Environment

Renewable energy technologies/sources path within EU 2020 strategy

Vattenfall Capital Markets Day 2008

Markets and costs of renewable energy

Climate Change and the Future Nordic Energy System

Summary Report UK s Long Term Power Balance and Implications for Major Power Producers Gas Consumption

Context Three numbers and three core global energy challenges: 6.5 million premature deaths each year can be attributed to air pollution 2.7 degrees i

Data, tables, statistics and maps ENERGY IN DENMARK

RENEWABLE ENERGY IN AUSTRALIA

Rana Adib Executive Secretary. GSR2018 Launch - Webinar Clean Energy Solutions Center 5 June 2018

Gasification of Biomass and Waste Recent Activities and Results of IEA Bioenergy Task 33

Renewable Energy: Technology, Economics and Environment

I. CITIES AND ENERGY USE

Introduction to medium term reports Based on the most recent data available 5 years outlook is important for policy making Natural gas and renewables

(How) can the European power market be decarbonized to 2050, without CCS? Lasse Torgersen CenSES årskonferanse Oslo 7/

EU Climate and Energy Policy Framework: EU Renewable Energy Policies

Data, tables, statistics and maps ENERGY IN DENMARK

Long-Term Policy: Concepts, Methods, Industry Practice

Vattenfall s views on the electricity market 2005

GHG emissions per capita. (tco 2. e/cap) Source: UNDP, data for 2015 Source: World Bank Indicators, data for 2012 Source: IEA, data for 2013

Portugal Energy efficiency report

ECONOMIC TRENDS OF ITALIAN ELECTRICITY SECTOR SHORT TERMS STATISTICS

The impacts of nuclear energy and renewables on network costs. Ron Cameron OECD Nuclear Energy Agency

Renewable Energy Sources in EU - Current status, future developments and challenges

Chapter 2. Comparison of Technologies. December 2015

GHG emissions per capita. (tco 2. e/cap) Source: UNDP, data for 2015 Source: World Bank Indicators, data for 2012 Source: IEA, data for 2013

Renewables 2O18 Analysis and Forecasts to 2O23

John Gale. General Manager 58 th WPFF Meeting Friendship Hotel, Beijing, China. 9 th to 10 th June 2010

Korea s Renewable Energy Policy - An Update -

Emissions Intensity CHAPTER 5 EMISSIONS INTENSITY 25

Table of contents. 1 Introduction System impacts of VRE deployment Technical flexibility assessment of case study regions...

Part 1: Sustainability in the Electricity Sector

Harness the wind to tackle climate change. = less CO 2

Supplementary Information

SOUTHEA ST ASIA ENERGY OUTLOOK EXECUTIVE SUMMARY. World Energy Outlook Special Report

Click to edit Master title style

STUDY SUMMARY (F) CDC-1319

138 ENVIRONMENTAL PROFILE OF SPAIN 2011

INTERNATIONAL ENERGY AGENCY. In support of the G8 Plan of Action TOWARD A CLEAN, CLEVER & COMPETITIVE ENERGY FUTURE

Consumption of energy

RENEWABLE ENERGY DEVELOPMENTS

Transcription:

International comparison of fossil power efficiency and CO2 intensity - Update 2017 FINAL REPORT

International comparison of fossil power efficiency and CO2 intensity Update 2017 FINAL REPORT By: Sam Nierop, Barry Vree and Viktorija Stojcheva Date: September 11, 2017 Project number: ESMNL17903 Ecofys 2017 by order of: MRI Research Associates, Japan ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

Summary The purpose of this study is to compare the power generating efficiency and CO 2-intensity of fossilfired power plants for Australia, China, France, Germany, India, Japan, Nordic countries (Denmark, Finland, Sweden and Norway aggregated), South Korea, United Kingdom and Ireland (aggregated), and the United States. This is done by calculating separate benchmark indicators for the generating efficiency of gas-, oil- and coal-fired power plants. Additionally, an overall benchmark for fossil-fired power generation is determined. The benchmark indicators are based on deviations from average generating efficiencies. The study is based on data from the IEA Energy Balances (2017 edition). Trends in power generation Total power generation in 2015 is largest in China with 5,649 TWh, with the United States following as second (4,119 TWh), and India (1,164 TWh) and Japan (827 TWh) as third and fourth respectively. Total power generation is smallest in Australia (227 TWh), United Kingdom and Ireland (314 TWh) and the Nordic countries (376 TWh). The countries included in the study generated 64% of public power generation worldwide in 2015, while the share of fossil power used in the public power production mix was 68%. Total fossil power generation is largest in China with roughly 4,178 TWh, exceeding the United States (2,758 TWh). India is the country ranked third with 944 TWh. From the fossil fuels, coal is most frequently used in all countries except Japan, France and UK & Ireland. Figure 1 shows the breakdown of public power generation per country. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Other renewables Oil Natural gas Coal Nuclear Hydro Figure 1 Public power generation by fuel source in 2015. Note that gas use in the Nordic countries is underestimated as Norwegian power production from natural gas is confidential. ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

Efficiency [%] Total coal-fired power generation in all countries in this study declined in 2015 compared to 2014 from 7,640 to 7,426 TWh (-3%), which is the first major decrease after multiple decades of growth. The United States in particular saw a major decrease of 14% compared to 2014 driven by relative low natural gas prices due to shale gas development Total gas-fired power generation in all countries in this study increased significantly from 1,939 in 2014 to 2,152 TWh in 2015 (+11%). The United States especially shows a strong absolute growth from 1,065 to 1,273 TWh (+20%), due to the abovementioned shale gas development. Oil-fired power generation played a marginal role in 2015 with only 116 TWh (around 1% of overall fossil-fired power generation). Japan and the United States were the largest oil-fired power producers and generated 83% of all oil-fired power production in the countries under consideration in this study. The general trend is that power production from oil has been declining over 1990 2014, although some temporary peaks can still be observed (e.g. Japan almost tripled power production from oil in 2012 compared to 2010 after the Fukushima incident). Generating efficiency Figure 2 shows the generating efficiency per country and fuel source according to the statistics used. Because the uncertainty in the efficiency for a single year can be high, we show the average efficiencies for the last three years available: 2013 2015. The following results can be highlighted: Coal-fired power efficiencies range from 34% (India) to 42% (Japan). Gas-fired power efficiencies range from 38% (India) to 58% (Korea). Oil-fired power efficiencies range from 26% (India) to 43% (Japan). Fossil-fired power efficiencies range from 34% (India) to 45% (Japan). The weighted average generating efficiency 1 for all countries together in 2015 is 38% for coal, 49% for natural gas, 40% for oil-fired power generation and 40% for fossil power in general. 60% 50% 40% 30% 20% 10% 0% Coal Gas Oil Fossil Figure 2 Generating efficiency per fuel source (average 2013 2015). 1 See section 2 for details on definitions and calculations. ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

Efficiency [%] The weighted average efficiency for gas-fired power generation shows a strong increase from 38% in 1990 to 49% in 2015 for the considered countries (see Figure 3). This is caused by a strong increase in modern gas-based capacity: gas-based production more than quadrupled. However, while coal-fired power generation more than doubled in the period 1990-2015, the weighted average efficiency only increased slightly from 35% to 38%. The reason for this is that best available technology is not applied widely, as the efficiency that can be achieved by applying best available technology (super-critical units) is as high as 47%. In particular in India, where a significant part of the growth in coal-fired power generation took place, generating efficiencies of coal remain below average. The weighted average efficiency for oil-fired power generation shows a peak in 2000 and 2001 due to a peak of the oil-fired generation efficiency in the United States, which is likely caused by data inconsistencies. 50% 45% 40% 35% Coal Gas Oil Fossil 30% 25% 1990 1995 2000 2005 2010 2015 Figure 3 Weighted average generating efficiency for included countries. Figure 4 shows the benchmark for the weighted efficiency of fossil-fired power generation for 2013-2015. Countries with benchmark indicators above 100% perform better than average and countries below 100% perform worse than the average. As can be seen, in order of performance, South Korea, United Kingdom and Ireland, Japan, the Nordic countries, Germany, China and France all perform better than the benchmark fossil-fired generating efficiency. United States, Australia and India perform below the average generating efficiency for fossil power. ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

CO 2 intensity [g/kwh] Performance relative to benchmark 120% 110% 100% 90% 80% 70% 60% 50% 40% Coal Gas Oil Fossil Weighted benchmark Figure 4 Benchmark for weighted generating efficiency of fossil-fired power plants for period 2013-2015 (100% is average). CO 2-intensity and reduction potential Figure 5 shows the CO 2-intensity for fossil-fired power generation for the years 2013 2015 per country. The CO 2 intensity for fossil-fired power generation ranges from 592 g/kwh for Japan to 975 g/kwh for India. The CO 2 intensity for fossil-fired power generation depends largely on the share of coal in fossil power generation and on the efficiency of power production. 1,200 1,000 800 600 400 200 0 2013 2014 2015 Average Figure 5 CO2-intensity for fossil-fired power generation for period 2013-2015. If the best available technologies (BAT) 2 would have been applied for all fossil power generation in the countries of this study in 2015, absolute emissions would have been, on average, 19% lower. Figure 6 shows how much lower CO 2 emissions would be for all individual countries as a share of emissions from 2 Installations operating according to the present highest existing conversion efficiencies. ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

CO 2 reduction potential [Mt CO 2 ] CO 2 reduction potential [%] fossil-fired power generation. The emission reduction potential per country ranges from 13% for Japan to 27% for India. 30% 25% 20% 15% 10% 5% 0% Figure 6 Relative CO2 emission reduction potential for fossil power generation by efficiency improvement by replacing all fossil public power production by BAT for the corresponding fuel type in 2015. Figure 7 shows the emission reduction potential in absolute amounts that could be achieved by replacing existing capacity with BAT. China, United States and India show very high absolute emission reduction potentials of 648, 387 and 243 Mt CO 2, respectively. This is mainly due to large amounts of coal-fired power generation at relatively low efficiency. 700 600 500 400 300 200 100 0 Figure 7 Absolute CO2 emission reduction potential for fossil power generation by efficiency improvement by replacing all fossil public power production by BAT for the corresponding fuel type in 2015. ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

Table of contents 1 Introduction 1 1.1 Power generation by fossil-fuel sources 1 2 Methodology 6 2.1 Efficiency of power generation 6 2.2 Benchmark for generating efficiency of fossil-fired power plants 7 2.3 CO 2 intensity power generation 10 2.4 Share of renewable and nuclear power generation 11 3 Results 12 3.1 Efficiency of coal-, gas- and oil-fired power generation 12 3.2 Benchmark based on non-weighted average efficiency 20 3.3 Benchmark based on weighted average efficiency 23 3.4 CO 2-intensities 27 3.5 Emission reduction potential 30 3.6 Renewable and nuclear power production 33 4 Conclusions 54 5 Discussion of uncertainties & recommendations for follow-up work 56 6 References 58 Appendix I: Generating efficiencies and comparison previous Ecofys report 60 Appendix II: Input data 65 Appendix III: IEA Definitions 79 ECOFYS Netherlands B.V. Kanaalweg 15G 3526 KL Utrecht T +31 (0)30 662-3300 F +31 (0)30 662-3301 E info@ecofys.com I www.ecofys.com Chamber of Commerce 30161191

1 Introduction This study is an update of the analysis International comparison of fossil power generation and CO 2 intensity (Ecofys, 2016). This analysis aims to compare fossil-fired power generating efficiency and CO 2-intensity (coal, oil and gas) for Australia, People's Republic of China, France, Germany, India, Japan, South Korea, Nordic countries (Denmark, Finland, Sweden and Norway aggregated), United Kingdom and Ireland, and the United States. This selection of countries and regions is based on discussions with the client. United Kingdom and Ireland, and the Nordic countries are aggregated, because of the interconnection between their electricity grids. Although the electricity grids in Europe are highly interconnected, there are a number of markets that operate fairly independently. These include the Nordic market (Denmark, Finland, Sweden and Norway), the Iberian market (Spain and Portugal) and United Kingdom and Ireland (European Commission, 2016). The analysis is based on the methodologies described in Phylipsen et al. (1998) and applied in Phylipsen et al. (2003). Only public power plants are taken into account, including public CHP plants. For the latter a correction for the (district) heat supply has been applied. This chapter gives an overview of the fuel mix for power generation for the included countries and of the amount of fossil-fired power generation. The methodology for this study is described in Chapter 2. Chapter 3 gives an overview of the efficiency and CO 2 intensity of fossil-fired power generation by fuel source and addresses the development of the share of renewables in public power generation over time. Chapter 4 outlines the main conclusions. Chapter 5 discusses uncertainties in data and analysis, and provides recommendations for detailed follow-up actions. 1.1 Power generation by fossil-fuel sources Fossil-fired power generation is a major source of greenhouse gas emissions. Worldwide, electricity and heat generation accounted for 42% of total greenhouse gas emissions in 2014 (IEA Emissions, 2016). The countries included in the study generated 64% of public power generation worldwide in 2015 (IEA, 2017). In 2015, the total power generation (incl. renewables and nuclear power) was largest in China with roughly 5,649 TWh, exceeding the United States (4,119 TWh), see Figure 8. Japan generated 827 TWh. The share of fossil fuels in the overall fuel mix for electricity generation was 68%. France, which has a large share of nuclear power (80%) and the Nordic countries with a large share of hydropower (59%) in 2015 are exceptions. It also notable that the share of other renewables is at 28% in Germany. When comparing the fossil fuel sources, Figure 8 shows that coal is most frequently used in all countries except for Japan, France and UK & Ireland where natural gas is more abundantly used. Australia, China and India show a very high share of coal in their overall fuel mix for power generation of around 70-1

Electricitry production (TWh) 77% in 2015. The share of oil-fired power generation is typically limited; only Japan, Korea and the United States have larger amounts, in absolute sense. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Other renewables Oil Natural gas Coal Nuclear Hydro 6,000 5,000 4,000 3,000 2,000 1,000 0 Other renewables Hydro Nuclear Oil Natural gas Coal Figure 8 Absolute (top) and relative (bottom) public power generation by source in 2015. Note that for the Nordic countries there is a small underestimation of power production from natural gas as power produced from natural gas in Norway is confidential. 2

Electricity generation (TWh) Figure 9 - Figure 12 show the amount of coal-, gas-, oil- and total fossil-fired power generation respectively in the period 1990-2015, from public power plants and public CHP plants together. The total coal-fired power generation in all countries increased from 3,043 to 7,640 TWh (+151%) during the period 1990 2014, but declined to 7,426 TWh in 2015 (-3% compared to 2014). The US saw its share of coal-fired power production shrink with 14% compared to 2014, mainly driven by national and regional regulations promoting (shale) gas and renewable technologies at the expense of coal-fired generation. 4500 4000 3500 3000 2500 2000 1500 1000 500 0 1990 1995 2000 2005 2010 2015 Australia China France Germany India Japan Korea Nordic countries UK + Ireland United States Figure 9 Coal-fired power generation from 1990 to 2015 for analysed countries Gas-fired power generation in all countries combined increased from 515 to 2,152 TWh in 1990 2015 (+318%), with an 11% increase in 2015 compared to 2014. Figure 10 shows that the United States especially showed a strong absolute growth from 1,065 TWh to 1,273 TWh (+20%) in 2015 compared to 2014, which was fuelled by shale gas development. In Japan, gas-fired power generation increased significantly post-fukushima (from 281 TWh in 2010 to 374 TWh in 2012). Oil-fired power generation plays a limited role and its importance has further diminished over time, especially in the case of the three leading oil-fired power producing countries (USA, Japan and South Korea). However, Figure 11 shows that Japan almost tripled power production from oil in 2012 compared to 2010 most likely due to the need for deploying reserve capacity as nuclear power plants were shut down after the Fukushima accident. 3

Electricity generation (TWh) Electricity generation (TWh) 1400 Australia 1200 China 1000 France 800 Germany India 600 Japan 400 Korea 200 Nordic countries UK + Ireland 0 1990 1995 2000 2005 2010 2015 United States Figure 10 Gas-fired power generation from 1990 to 2015 for analysed countries 250 Australia 200 China France 150 Germany India 100 Japan Korea 50 Nordic countries UK + Ireland 0 1990 1995 2000 2005 2010 2015 United States Figure 11 Oil-fired power generation from 1990 to 2015 for analysed countries 4

Electricity generation (TWh) Figure 12 indicates that China, US and India had the strongest absolute growth in fossil-fired power production from 1990 to 2015. The total fossil-fired power generation increased steadily from 4,000 in 1990 to 9,718 TWh (+143%) in 2013, but has stabilized at that level in 2014 and 2015. Indeed, of the countries in this study only India has significantly increased its fossil-fired power production from 2013 to 2015. Most European countries have decreased their fossil-fired power production over this two-year period (-25% for France, -8% for Germany, -42% for Nordic countries and -22% for UK & Ireland). 4500 4000 3500 3000 2500 2000 1500 1000 500 0 1990 1995 2000 2005 2010 2015 Australia China France Germany India Japan Korea Nordic countries UK + Ireland United States Figure 12 Fossil-fired power production from 1990 to 2015 for analysed countries 5

2 Methodology This chapter discusses the methodology used to derive the generating efficiency indicators as well as the input data used to determine the indicators. This study is based on data from IEA World Energy Balances edition 2017 (IEA, 2017). The advantage of using IEA Energy Balances is its consistency on a number of points: Energy inputs for power plants are based on net calorific value (NCV) 3 ; The output of the electricity plants is measured as gross production of electricity and heat. This is defined as the electricity production including the auxiliary electricity consumption and losses in transformers at the power station ; A distinction is made between electricity production from industrial power plants and public power plants and public combined heat and power (CHP) plants. In this study we take into account public power plants and public CHP plants. We distinguish three types of fossil fuel sources: (1) coal and coal products, (2) crude oil and petroleum products and (3) natural gas. In the remainder of this report, we will refer to these fuel sources as coal, oil and gas, respectively. For a more extensive definition of public power production and these fuel types, refer to Appendix III: IEA Definitions. 2.1 Efficiency of power generation The formula for calculating the efficiency of power generation is: E = (P + H*s) / I. Where: E P H s I Efficiency of power generation Power production from public power plants and public CHP plants Heat output from public CHP plants Correction factor between heat and electricity, defined as the theoretical reduction in electricity production per unit of heat produced Fuel input for public power plants and public CHP plants Heat extraction causes the efficiency of electricity generation to decrease although the overall efficiency for heat and electricity production is higher than when the two are generated separately. Therefore, a correction for heat extraction is applied. This correction reflects the amount of electricity production lost per unit of heat extracted from the power plant(s). For district heating 3 The Net Calorific Value (NCV) or Lower Heating Value (LHV) refers to the quantity of heat liberated by the complete combustion of a unit of fuel when the water produced is assumed to remain as a vapour and the heat is not recovered. 6

systems, the substitution factors vary between 0.15 and 0.2. In our analysis we have used a value of 0.175. It must be noted that when heat is delivered at higher temperatures (e.g. to industrial processes), the substitution factor can be higher. However, at the moment, the amount of hightemperature heat delivered to industry by public utilities is small in most countries. We estimate that the effect on the average efficiency is not more than an increase of 0.5 percent point 4. No corrections are applied for air temperature and cooling method. The efficiency of power plants is influenced by the temperature of the air or cooling water. In general surface water-cooling leads to higher plant efficiency than the use of cooling towers. The cooling methods that can be applied depend on local circumstances, like the availability of abundant surface water and existing regulations. The effect of cooling method on efficiency may be up to 1-2 percent points. Furthermore the efficiency of the power plant is affected by the temperature of the cooling medium. The sensitivity to temperature can be in the order of 0.1-0.2 percent point per degree (Phylipsen et al, 1998). In order to determine the efficiency for power production for a region, we calculate the weighted average efficiency of the countries included in the region. 2.2 Benchmark for generating efficiency of fossil-fired power plants In this analysis we compare the generating efficiency of fossil-fired power plants across countries and regions. Instead of simply aggregating the efficiencies for different fuel types to a single efficiency indicator, we determine separate benchmark indicators per fuel source. This is because the efficiency for natural gas-fired power generation is generally higher than the efficiency for coalfired power generation. In general, choices for fuel types are often outside the realm of the industry and therefore a structural factor. Choices for fuel diversification have in the past often been made at the government level for strategic purposes, e.g. fuel diversification and fuel costs. The most widely used power plants for coal-fired power generation are conventional boiler plants based on the Rankine cycle. Fuel is combusted in a boiler and with the generated heat, pressurized water is heated to steam. The steam drives a turbine and generates electricity. In principle any fuel can be used in this kind of plant. An alternative for the steam cycle is the gas turbine, where combusted gas expands through a turbine and drives a generator. The hot exit gas from the turbine still has significant amounts of energy which can be used to raise steam to drive a steam-turbine and another generator. This combination of gas and steam cycle is called combined cycle gas turbine (CCGT) plant. A CCGT plant is generally fired with natural gas. Also coal firing and biomass firing is possible by gasification, e.g. in integrated coal gasification combined cycle plants (IGCC). These technologies 4 A change of 1 percent point in efficiency here means a change of e.g. 40% to 41%. 7

are not widely used. The generating efficiency of a single steam cycle is at most 47%, while the generating efficiency of a combined cycle can be over 61% (Siemens. 2012; GE 2014 5 ). Open cycle gas turbine plants (only gas turbine, no steam cycle) are also still widely applied. Several possible indicators exist for benchmarking efficiency of power generation. One possible indicator is the comparison of individual countries efficiencies to predefined best practice efficiency. The difficulty in this method is the definition of best practice efficiency. Best practice efficiency could e.g. be based on: The best performing country in the world or in a region; The best performing plant in the world or in a region; The best practical efficiency possible, by best available technology (BAT). The best practice efficiency differs yearly, which means that back-casting is required to determine best practice efficiencies for historic years. A different method for benchmarking generating efficiency is the comparison of countries efficiencies against average efficiencies. An advantage of this method is the visibility of a countries performance against average efficiency. In this study we choose to use this indicator. We compare the efficiency of countries and regions to the average efficiency of the selected countries. The average efficiency is calculated per fuel source and per year and can be either weighted or non-weighted. In the first case the weighted-average efficiency represents the overall generating efficiency of the included countries. A disadvantage of this method is that countries with a large installed generating capacity heavily influence the average efficiency while small countries have hardly any influence at all on the average efficiency. On the other hand, when applying nonweighted benchmark indicators, one efficient power plant in a country could influence the average efficiency if absolute power generation in the country is small. In this research we included both methods, to verify if this leads to different results. The formula for the non-weighted average efficiency for coal (BC 1) is given below as an example. The formulas for oil and gas are similar. 5 https://powergen.gepower.com/plan-build/products/gas-turbines/9ha-gas-turbine.html 8

BC 1 = EC i / n Where: BC 1 EC i n Benchmark efficiency coal (1). This is the average efficiency of coal-fired power generation for the selected countries. Efficiency coal for country or region i (i = 1, n) The number of countries and regions The formula for the weighted average efficiency for coal (BC 2) is given below as an example: BC 2 = (PC i + HC i *s)/ IC i Where: BC 2 PC i HC i s IC i Benchmark efficiency coal (2). This is the weighted average efficiency of coal-fired power generation for the selected countries. Coal-fired power production for country or region i (i = 1, n) Heat output for country or region i (i = 1, n) Correction factor between heat and electricity, defined as the reduction in electricity production per unit of heat extracted Fuel input for coal-fired power plants for country or region i (i = 1, n) To determine the performance of a country relative to the benchmark efficiency we divide the efficiency of a country for a certain year by the benchmark efficiency in the same year. The formula of the indicator for the efficiency of coal-fired power is given below as an example: BC i = EC i / BC 1 or BC i = EC i / BC 2 Where: BC i Benchmark indicator of coal-fired generating efficiency for country or region i Countries that perform better than average for a certain year show numbers above 100% and vice versa. To come to an overall comparison for fossil-fired power efficiency we calculate the outputweighted average of the three indicators, as is shown in the formula below: BF i = (BC i * PC i + BG i * PG i + BO i * PO i) / (PC i + PG i + PO i) Where: BF i, BC i, BG i and BO i PC i, PG i and PO i Benchmark indicator for the efficiency of fossil-fired, coal-fired, gasfired and oil-fired power generation for country or region i Coal-fired, gas-fired and oil-fired power production for country or region i 9

2.3 CO2 intensity power generation In this study we also calculate CO 2 emissions intensities per country: Per fossil fuel source (coal, oil, gas); For total fossil power generation and For total power generation. There are several ways of calculating CO 2-intensities (g CO 2/kWh) for power generation, depending on the way combined heat and power generation is taken into account. In this study we use the same method as for calculating overall generating efficiency and correct for heat generation by the correction factor of 0.175 (see Section 2.1). The formula for calculating CO 2 intensity is: CO 2-intensity = (3.6 * C i * P i /E i) / P i Where: i Fuel source 1... n E i Efficiency power generation per fuel source (see Section 2.1) C i P i CO 2 emission factor per fuel source (see table below) (tonne CO 2/TJ) Power production from public power and CHP plants per fuel source (GWh) The table below gives the CO 2 emission factors per fuel source. Table 1 Fossil CO2 emission factor (IEA, 2005) Fuel type Tonne CO2/TJncv Hard coal 94.6 Lignite 101.2 Natural gas 56.1 Oil 74.1 Other fuels (biomass, nuclear, etc.) 0 10

2.4 Share of renewable and nuclear power generation This report also gives an insight into the development of the share of renewable and nuclear power production in total public power production. For the period 2000-2015, annual developments for all geographical regions as stated above are included. The IEA classifies a number of different energy sources that are used for power production as renewables (see Table 2). Ecofys has mapped (i.e. aggregated) these into various different categories: Bio; Geothermal; Hydro; Solar; Ocean; Waste; Wind. Table 2 Mapping of different renewable energy categories of IEA Renewable energy sources as defined by IEA Industrial waste Municipal waste (renewable) Primary solid biofuels Biogases Bio-gasoline Biodiesels Other liquid biofuels Non-specified primary biofuels and waste Charcoal Hydro Geothermal Solar photovoltaics Solar thermal Tide, wave and ocean Wind Ecofys mapping Waste Waste Bio Bio Bio Bio Bio Bio Bio Hydro Geothermal Solar Solar Ocean Wind Data input for calculating the shares originates from IEA, 2017. To be consistent with the rest of this study, only the share in public power production is considered, omitting the installed capacity in the private sector (i.e. energy production for own use). 11

3 Results Table 3 gives an overview of the content of the different sections of Chapter 3. Table 3 What can be found in which section in this chapter Section 3.1 3.2 Content Efficiencies for coal,- gas- and oil-fired power production, including a simple aggregation of fossil-fired power efficiency Results of the benchmark analysis based on non-weighted average efficiencies 3.3 Results of the benchmark analysis, based on weighted average efficiencies 3.4 CO 2 intensities per fuel source and for total power generation per country 3.5 3.6 CO 2 abatement potentials per country when replacing current installed generating capacity by best available technology (BAT) Development of the share of renewable and nuclear power production over the last decade The underlying data for the figures in this chapter can be found in Appendix II: Input data, which gives the input for the analysis in terms of power generation, fuel input, heat output, benchmark efficiencies and CO 2-intensity. 3.1 Efficiency of coal-, gas- and oil-fired power generation Figure 13 - Figure 15 show the efficiency trend for coal-, gas- and oil-fired power production, respectively, for the period 1990-2015. Figure 16 shows the efficiency of fossil-fired power generation by the weighted-average efficiency of gas, oil- and coal-fired power generation. 12

Efficiency [%] 45% Australia 40% China France 35% Germany India 30% Japan 25% Korea Nordic countries 20% 1990 1995 2000 2005 2010 2015 UK + Ireland United States Figure 13 Average efficiency of coal-fired power production from 1990 to 2015 for analysed countries The efficiencies for coal-fired power generation in 2015 range from 35% for India to 42% for Japan. Note that over the last five years China and India, two of the largest coal-fired power producers, have continued to improve the efficiency of their coal-fired power generation. Most of the other countries have experienced very limited improvement or efficiency decreases. 13

60% Australia 55% China 50% France Efficiency [%] 45% 40% 35% 30% 25% 20% 1990 1995 2000 2005 2010 2015 Germany India Japan Korea Nordic countries UK + Ireland United States Figure 14 Average efficiency of gas-fired power production from 1990 to 2015 for analysed countries In contrast to other fuels, the efficiency of gas-fired power generation has improved substantially over the last two decades. The largest efficiency improvements in 1990-2015 are observed for India 6, Germany and South Korea. In 2015, efficiencies range from 38% for Australia to 58% for South Korea. The increase in efficiency for South Korea in the last years might be because of the recent construction of some highly efficient combined cycle power plants. Generating efficiencies for gas power plants in India appear to have some inconsistencies in the early 2000s, as average efficiencies of around 58% are close to BAT and thus can be considered unrealistic for India at that time. The sudden peak for Australia in 2003 and 2004 is also noteworthy and might indicate inconsistencies in the statistics. For some countries, such as France, efficiencies fluctuate heavily over time. This may be explained by gas-fired power plants significantly varying operating hours from year to year. Fourfifths of the French public power originates from nuclear plants, with natural gas only responsible for 3% in 2014. Natural gas capacity is deployed as a peak load capacity together with oil fired capacity. This provides explanation for the fluctuating and relatively low efficiencies. 6 Although the early figures for India are deemed unreliable with efficiencies approaching a BAT efficiency of 61% in 1999. 14

Efficiency [%] 60% Australia 50% China France 40% Germany 30% India Japan 20% Korea Nordic countries 10% 1990 1995 2000 2005 2010 2015 UK + Ireland United States Figure 15 Average efficiency of oil-fired power production from 1990 to 2015 for analysed countries In 2015, the efficiencies for oil-fired power generation range from 26% for India to 43% for Australia. The graph shows large fluctuations in efficiency for oil-fired power generation, e.g. for France efficiencies peak above BAT levels. The explanation for the variance could be partly the fluctuation of yearly operating hours, as running at significantly lower operating hours typically lowers efficiencies, or another explanation could be data uncertainty in the case of unrealistically large changes. It should be noted that oil-fired power generation is relatively small (below 3 TWh in 2015) in all countries but Japan, South Korea and the USA. Therefore, the overall impact on the average fossil-fired generating efficiency is limited. 15

Efficiency [%] 50% Australia 45% 40% China France Germany 35% India 30% 25% Japan Korea Nordic countries 20% 1990 1995 2000 2005 2010 2015 UK + Ireland United States Figure 16 Average efficiency of fossil-fired power production from 1990 to 2015 for analysed countries For overall fossil-fired power generation, the efficiencies range from 35% for India to 45% for Japan in 2015. Below is a discussion of the results organised by country. Note that all data refers to the year 2015, unless stated otherwise. Australia Total fossil-fired power generation in Australia is 201 TWh, of which 79% is generated from coal. Gas-fired power generation represents most the remainder of fossil-fired power generation (20%), with oil-fired power generation in Australia only accounting for 1%. The efficiency of gas-fired power generation in Australia was very low at 38% in 2015. This makes Australia now the lowest performer in gas-fired power generation efficiencies, taking over this spot from India, which was the bottom performer in previous years. The efficiency for coal-fired power generation of Australia also ranks the near the bottom at 35.2%, which is close to the lowest performer (India at 34.6%). However, Australia is surprisingly the top performer for oilfired power generation efficiencies at 43%. 16

China China is the largest fossil-fired power generator, generating 4,178 TWh in 2015. China relies almost entirely on coal (96% of fossil power production). Gas-fired power generation increased significantly from 115 TWh in 2014 to 145 TWh in 2015 (+27%) and now comprises 3% of fossilfired power generation. Oil-fired power generation is negligible in China in 2015 (roughly 1 TWh). The efficiency of coal-fired power generation is 38.6%. It has increased steadily in the period 1990-2014 coming from 28.8%. Coal-based electricity production increased substantially from 440 TWh in 1990 to 4,025 TWh in 2013, corresponding to about a tenfold increase. However, the last two years coal-based electricity production has stabilized at this level (4,040 TWh in 2014 and 4,032 TWh in 2015). This is an important reason that global coal-fired power production in 2015 decreased significantly compared to 2014, which is the first major decrease since decades. France Fossil-fired power generation in France was relatively small at only 28 TWh in 2015. Besides the dominant contribution of nuclear capacity, France fossil fuel use is mainly coal and gas-fired power generation. Coal-fired power generation in France shows strong fluctuations ranging from about 20 to 50 TWh per year in the past two decades, which can be explained by the existence of power production that has lower marginal production costs, i.e. hydro and nuclear plants. Fossil power generation is used to absorb electricity demand peaks in winter due to electric heating, which means that the capacity factor of coal-fired power plants can vary strongly year-by-year. The generating efficiency for coal-fired power plants in France was 40.2% in 2015. Public gas-fired power generation increased from practically zero in 1990 to roughly 22 TWh in 2011, decreased to 8 TWh in 2014, but has picked up again to 15 TWh in 2015. The generating efficiency of natural gas is increasing recently (from 43% in 2012 to 48% in 2015) because of the commissioning of highly efficient combined cycle plants. Germany Fossil-fired power generation in Germany totalled 309 TWh in 2015, of which 87% is produced by coal. After the reunification of West and East Germany several inefficient lignite power plants were closed. This led to a higher efficiency of coal-fired power generation, which increased gradually from 34.4% in 1990 to 39.6% in 2015. In the mid '90s the natural gas market was liberalised in Germany, leading to more competition and lower gas prices. This resulted in more gas use and a large increase of CHP capacity. This led to a strong increase of efficiency of gas-based power generation from 32.6% in 1990 to 47.1% in 2015. Gas-fired power generation increased from 25 TWh in 1990 to 72 TWh in 2008, but has been gradually decreasing since 2008, standing at 39 TWh in 2015. 17

India Fossil-fired power generation in India is 944 TWh, of which 95% is produced from coal. Gas-fired power generation increased from 8 TWh in 1990 to 98 TWh in 2010, but recently decreased significantly, standing at 45 TWh in 2015. Oil-fired power generation was 3 TWh in 2015. The efficiency for coal-fired power generation is constantly quite low around 31-33% over the whole period of 1990-2010. Some reasons for this may be (IEA, 2003): The coal is unwashed; Indian coal has a high ash content of 30% to 55%; Coal-fired capacity is used for peak load power generation as well as base load power generation. However, in recent years since 2010 the efficiency for coal-fired power generation has been steadily increasing significantly, and with an efficiency of 34.6% India is now approaching the second-lowest performer (Australia at 35.2%). The efficiency for gas-fired power generation increased from 23.2% in 1990 to 40.0% in 2014, suggesting strong efficiency developments. However, it should be noted that the efficiency is highly fluctuating (e.g. 57.7% in 1999 to 35.7% in in 2010). Although efficiencies in India have significantly improved because of the installation of modern combined cycle plants (IEA, 2003), the statistics should be taken with care because of inconsistencies in the early 2000 s where the efficiency seems unrealistically high. Japan Japan is the fourth largest fossil-fired power producer of the countries studies with 732 TWh, starting at 466 TWh in 1990. Fossil-fired power production increased significantly in 2011 to 731 TWh (from 611 TWh in 2010, +20%) to compensate for the shutdown of nuclear power plants following the Fukushima incident. After peaking at 809 TWh in 2013, fossil-fired power generation has been decreasing the past two year to 732 TWh in 2015. The largest part of this decrease comes from a diminishing oil-based production (from 129 TWh in 2013 to 66 TWh in 2015). The efficiency for coal-fired power generation slowly increased from 39.8% in 1990 to 42.9% in 2010, but decreased slightly to 42.1% in 2015. Similarly, the efficiency for oil-fired power generation slowly increased from 40.8% in 1990 to 44.2% in 2013, but decreased slightly to 42.4% in 2015. However, the gas-fired generating efficiency in Japan is still increasing, as it reached the highest point ever in 2015 (49.2%). The Japanese Central Research Institute of the Electric Power Industry (CRIEPI) describes the development of gas-fired power plants in Japan as follows: Japanese general electric utilities started to implement gas-fired power plants ahead of time in response to the oil crises of the 1970s. In those times gas turbines were not yet implemented on a large scale. As a result, utilities implemented conventional steam turbines based on active electricity demand, as they remain now. In the 1990s however, utilities 18

implemented combined cycle power plants. Furthermore, utilities will implement More Advanced Combined Cycle (MACC) with 59% (LHV) thermal efficiency, among the world s highest. The first MACC began its commercial operation in June 2007. Nordic countries Total fossil-fired power generation in the Nordic countries was 23 TWh in 2015. Finland and Denmark comprise the majority of this generation with 55% and 38% respectively. Sweden accounts for only 6% and Norway for 1%. However, power production in Norway might be underestimated as power produced from natural gas in Norway is confidential data. All types of fossil-fired power generation in the Nordic countries has been decreasing in recent years. Coal-fired power generation reached its lowest point since 1990 at 16 TWh in 2015. The efficiency for coal-fired power generation in the Nordic countries has been between 39.5% and 41.7% in 1990 2014, but it decreased below this range in 2015 with 38.9%, likely because of lower operating hours. Gas-fired power generation (with a large share consisting of CHP plants) in the Nordic countries is 6 TWh, generated at an efficiency of 48.1%. South Korea Total fossil-fired power generation in South Korea was 344 TWh in 2015, of which 63% is generated by coal and 35% by gas. South Korea has also significant oil-fired power production at 9 TWh in 2015, even though this has been decreasing in recent years. The efficiency for coal-fired power generation increased strongly from 25.8% in 1990 to 36.5% in 1995 and has fluctuated significantly since then between 33% and 39%. The efficiency of gasfired power generation increased from 40.5% in 1990 to 50% in 2002 and remaining at these levels until 2011. Recently, the efficiency of gas-fired power generation increased significantly to 58% in 2015. This increase might be because of the recent construction of some highly efficient combined cycle power plants. United Kingdom and Ireland Total fossil-fired power generation in the United Kingdom and Ireland was 178 TWh, of which 47% is generated from coal and 53% from gas. As gas prices decreased during the liberalization in the 90 s, gas-fired power generation capacity increased significantly from 1992 onwards, from only 4 TWh in 1990 to 170 TWh in 2010. However, gas-fired power generation has dropped significantly in recent years falling to 94 TWh in 2015. Coal-fired power generation has also decreased significantly from 148 TWh in 2012 to 83 TWh in 2015. The generating efficiency for coal-fired power plants has remained relatively constant over the past 20 years around 37-39%, it was 38.9% in 2015. This relative constant efficiency can be explained by the fact that no renewal of the coal-based stock has occurred. For natural gas the 19

trend is different: the large addition of new capacity has resulted in a strong increase of the average efficiency of gas-fired power plants, from 40.4% in 1990 to 52.8% in 2015. United States The United States is the second largest fossil-fired power generating country in the world with 2,758 TWh in 2015, of which 53% is generated by coal and 46% by gas. Coal-fired power generation decreased significantly from 1,695 TWh in 2014 to 1,455 in 2015 (-14%), which is the lowest point ever since 1990. Conversely, electricity generation by gas-fired power plants increased strongly in the past years from 1,065 TWh in 2014 to 1,273 TWh in 2015 driven by the availability and relative low prices of natural gas. The generating efficiency of coal-fired power generation remained to a high degree constant since 2002 at around 37%. The efficiency of gas-fired power generation has been gradually increasing from 37.2% in 1990 to 49.1% in 2015. 3.2 Benchmark based on non-weighted average efficiency In this section, a benchmark indicator for fossil-fired power generating efficiency is calculated. This is done by comparing the efficiency of countries and regions to the average efficiency of the selected countries. Separate benchmark indicators per fuel for coal, oil, gas and for total fossilfired power generation are calculated to compare the efficiencies. The formula for calculating the benchmark indicators can be found in Chapter 2. The benchmark indicator is based on the country efficiency per fuel source divided by the average efficiency per fuel source. The separate benchmark indicators are weighted by power generation to get to an overall indicator for fossilfired power generation. 20

Efficiency [%] Figure 17 shows the average efficiencies for all countries and regions considered in this study. Because these efficiencies are not weighted, they do not represent the total overall generating efficiency of power production in the included countries; one efficient power plant in a country could influence the average efficiency if absolute power generation in the country is small (see Section 2.2). 50% 45% 40% 35% Coal Gas Oil Fossil 30% 25% 1990 1995 2000 2005 2010 2015 Figure 17 Average non-weighted efficiencies With regard to average non-weighted efficiencies, the efficiency for gas-fired power generation shows a strong increase from 38% in 1990 to 48% in 2015 (average annual improvement of 1.0%). The reason for this improvement is mainly the large amount of new, more efficient generating capacity; gas-fired power generation more than quadrupled over the period 1990-2015. The increase in efficiency for coal-fired power generation was more limited, from 35% to 38% (average annual improvement of 0.4%). This is because coal-fired power generation increased relatively less (twofold increase over the period 1990 2015) and a significant part of the growth in coalfired power generation took place in India, where generating efficiency by coal is still significantly below BAT levels. 21

Performance relative to benchmark Figure 18 shows the generating efficiencies of the countries divided by the non-weighted average of efficiency. The data is averaged over the period 2013 2015 as uncertainty in the data of an individual year can be high. A benchmark indicator of 110% for gas means that the efficiency for gas-fired power generation in a country is 10% higher than the average (non-weighted) efficiency of the considered countries. The fossil benchmark indicator is based on the average benchmark indicators for coal, gas and oil, and is weighted by power generation output. 120% 110% 100% 90% 80% 70% 60% 50% Coal Gas Oil Fossil Non-weighted benchmark Figure 18 Average 2013 2015 performance for coal, gas, oil and fossil for countries relative to respective nonweighted average benchmark efficiencies. Countries are sorted on the basis of performance relative to the nonweighted benchmark for fossil fuel-fired power generation. As can be seen, Japan, the UK & Ireland and Korea perform best in terms of fossil-fired power generating efficiency with 11%, 9% and 6% above average efficiency respectively, followed by France with 4% above average. India and Australia are the most prominent underperformers with generating efficiencies at 15% and 12% below the benchmark respectively. Figure 19 shows the time development of the benchmark indicator for fossil-fired power generation. Note that a decrease of the benchmark indicator for a country might mean that the efficiency of the power production in the country has decreased or that the non-weighted average efficiency has increased. 22

Performance relative to fossil benchmark Australia 120% China 110% France Germany 100% India 90% Japan Korea 80% Nordic countries 70% 1990 1995 2000 2005 2010 2015 UK + Ireland United States Figure 19 Output-weighted average benchmark for generating efficiency of fossil-fired power plants (based on non-weighted average efficiencies). 3.3 Benchmark based on weighted average efficiency In this section, we calculate a second benchmark indicator for fossil-fired power generating efficiency. This is done by comparing the efficiency of countries and regions to the weighted average efficiencies of the selected countries. The formula for calculating the benchmark indicators can be found in Section 2.2. The benchmark indicator is based on the country efficiency per fuel source divided by the weighted average efficiency per fuel source. The separate benchmark indicators are weighted by power generation to get to an overall indicator for fossil-fired power generation. 23

Efficiency [%] Figure 20 shows the weighted average efficiencies for all countries and regions considered in this study. The weighted average generating efficiency for all countries and regions together in 2015 is 38% for coal, 49% for natural gas, 40% for oil-fired power generation and 40% for fossil power in general. 50% 45% 40% 35% Coal Gas Oil Fossil 30% 25% 1990 1995 2000 2005 2010 2015 Figure 20 Average weighted efficiencies of all countries and regions at the scope of this study (%). For the weighted average efficiencies, the efficiency for gas-fired power generation shows a strong increase from 38% in 1990 to 49% in 2014 (average annual improvement of 1.1%). The reason for this improvement is mainly the large amount of (more efficient) new generating capacity (see previous section). The efficiency for oil fired generation decreased in 2014 and 2015 compared to years before. This can be explained by a decrease in efficiency for the largest oil-fired power generators, the United States and Japan. The peak in 2000 and 2001 for the oil-based generation efficiency is due to efficiency peaks for the United States in these years, which is likely due to data inconsistencies. Coal-fired power generation increased more than twofold over the period 1990-2014. However, only a limited increase in efficiency is seen of 35% to 38% (average annual improvement of 0.3%). The reason for this is that a significant part of the growth in coal-fired power generation took place in India, where generating efficiency by coal is still significantly below BAT levels. The differences with the non-weighted average approach (Figure 17) are significant. In general these can be explained by the fact that the impact of countries with large power production output is diminished by the non-weighted approach whereas the impact of small countries is magnified. For instance, Korea and the United States are among the largest producers from gas-fired power and also among the most efficient. In the non-weighted average approach their impact on the average is lower than in the weighted average approach, resulting in a lower overall average efficiency for all countries combined. 24